Elastic High Performance Computing:Create accelerated image

Go to the Elastic Compute Service (ECS) instance buy page, select the Custom Launch tab, and meet the following requirements. For more information, see Custom purchase of an instance.

• Billing Method: Select Pay-as-you-go.

• Instance: Select X86_64 architecture.

• Image: Select a CentOS 7 series image.

• Public IP Address: Select Assign Public IPv4 Address.

• Security Group: Select the security group used by the cluster.

Click Create Order to complete the instance creation.

After the instance status shows Running, connect to the created ECS compute node instance. For more information, see Use Workbench to connect to a Linux instance over SSH.

Download and install the Python environment, including Python 2 and 3 environments. In this example, Python 2.7 and Python 3.9 are used.

1. In the root directory, create an installation script named install_python.sh.

   #!/bin/bash
   
   PYTHON_INSTALL_ROOT=/usr/local/py3.9env
   SERVICE_INSTALL_ROOT=/usr/local/ehpc_service
   
   mkdir $PYTHON_INSTALL_ROOT
   
   set -e
   
   arch=`arch`
   srcEnvPath=$1  # Pass this as an argument when calling the script
   #py3=$2  # Pass this as an argument (0 or 1) to decide which Python version to install
   
   pythonenv="python3.9env"
   echo "Downloading and extracting Python 3.9.10 for ${arch}..."
   wget -c -nv ${srcEnvPath}/python3.9.10.$arch.tgz -O - | tar -xz -C /usr/local/py3.9env/
   echo "Downloading and extracting ${pythonenv} for ${arch}..."
   curl -s -O ${srcEnvPath}/${pythonenv}.$arch.tgz
   if [ -d /usr/local/py3.9env/${pythonenv} ]; then
     rm -rf /usr/local/py3.9env/${pythonenv}
   fi
   tar -xzf ${pythonenv}.$arch.tgz -C /usr/local/py3.9env/
   
   echo "Downloading and extracting Python 2.7 environment..."
   curl -s -O ${srcEnvPath}/centos/7.2/python2.7env.tgz
   if [ -d /usr/local/python2.7env ]; then
     rm -rf /usr/local/python2.7env
   fi
   tar -xzf python2.7env.tgz -C /usr/local/

2. Add execution permissions to the script.

   chmod +x install_python.sh

3. Execute the script to download and install.

   bash /root/install_python.sh http://public-ehs.vpc100-oss-cn-hangzhou.aliyuncs.com/packages/env

4. Check that the py3.9env and python2.7env directories exist in the /usr/local/ directory to confirm successful installation.

Download and install Node.js and other dependency packages.

1. In the root directory, create an installation script named install_lib.sh.

   #!/bin/bash
   
   set -e
   arch=`arch`  # Pass this as an argument when calling the script
   srcEnvPath=$1  # Pass this as an argument when calling the script
   
   # echo "Installing Node.js for aarch64..."
   # curl -s -O $srcEnvPath/node-v16.17.1-linux-arm64.tar.xz
   # tar -xf node-v16.17.1-linux-arm64.tar.xz
   # mv node-v16.17.1-linux-arm64 /usr/local/node-v16.17.1-linux-arm64
   # ln -sfT /usr/local/node-v16.17.1-linux-arm64/bin/node /usr/local/bin/node
   # ln -sfT /usr/local/node-v16.17.1-linux-arm64/bin/npm /usr/local/bin/npm
   
   echo "Installing Node.js for x86_64..."
   curl -s -O $srcEnvPath/node-v6.11.1-linux-x64.tar.gz
   tar -xzf node-v6.11.1-linux-x64.tar.gz
   mv node-v6.11.1-linux-x64 /usr/local/node-v6.11.1
   ln -sfT /usr/local/node-v6.11.1/bin/node /usr/local/bin/node
   ln -sfT /usr/local/node-v6.11.1/bin/npm /usr/local/bin/npm
   
   echo "Installing additional libraries..."
   yum clean all && yum makecache
   yum install -y perf postgresql-contrib hwloc-devel jsoncpp-devel libcurl-devel nfs-utils cifs-utils environment-modules unzip jq
   
   echo "Appending module path..."
   echo -e "\nexport MODULEPATH=/opt/ehpcmodulefiles" >> /etc/bashrc
   
   kernel_version="$(uname -r)"
   kernel_devel_pkg="kernel-devel-${kernel_version}"
   base_os="7"
   kernel_devel_remote_repo="http://mirrors.cloud.aliyuncs.com/opsx/ecs/linux/rpm/driver/${base_os}/x86_64"
   
   rpm -q "${kernel_devel_pkg}" || yum install -y "${kernel_devel_pkg}"
   
   
   echo "Installing xdragon-health-protector..."
   xdragon_pkg="xdragon-health-protector-vm-1.0.x86_64.rpm"
   curl -s -O $srcEnvPath/${xdragon_pkg}
   yum install -y ${xdragon_pkg}
   systemctl enable xdragon-health-protector
   systemctl restart xdragon-health-protector
   systemctl status xdragon-health-protector

2. Add execution permissions to the script.

   chmod +x install_lib.sh

3. Execute the installation script. Run the following command. If the status shows running, the installation is successful.

   bash /root/install_libs.sh http://public-ehs.vpc100-oss-cn-hangzhou.aliyuncs.com/packages/env 

   

Visit the ECS console - Instances page. Click the target instance to access the Instance Details page.

Click Create Custom Image to complete the creation of the accelerated image.

Log on to the E-HPC console. Click the target cluster name to access the Cluster Details page.

In the left navigation bar, select Nodes and Queues > Queue, and click Edit in the Actions column of the target queue.

On the Edit Queue page, complete the parameter configuration for Instance Type Group. For more information, see Configure auto scaling of nodes.

1. Click Add Instance to select the node specifications.

2. Click Custom Image to select the created accelerated image.

   

3. Click OK.

Click Save to complete the cluster auto scaling configuration.